Obesity Induces DNA Damage in Mammary Epithelial Cells Which Is Exacerbated by Acrylamide Treatment Through CYP2E1-Mediated Oxidative Stress

1. Introduction

Breast cancer is the second most diagnosed cancer in women [1]. In the United States, approximately one in eight women will receive a breast cancer diagnosis in her lifetime [2]. Family history of breast cancer, sex, age at first menarche, and age at menopause are all strong risk factors for breast cancer [3,4]. Additionally, there are modifiable lifestyle factors that contribute to breast cancer risk, including lack of physical activity, obesity, and chemical exposures such as alcohol, hormone replacement therapy, and carcinogens [3,4]. Individual risk factors may interact with each other to further amplify breast cancer risk. For instance, obesity increases breast cancer risk in individuals with other risk factors such as familial history of breast cancer [5] or ethnicity [6]. Rates of obesity have tripled since 1970 [7] and chemical production has increased fifty-fold since 1950 [8], indicating a rise in modifiable risk factor exposure and highlighting a need to understand how these risk factors impact the breast to promote tumorigenesis.

Women with obesity are at risk for different subtypes of breast cancer depending on whether they have reached menopause. Premenopausal women with obesity are at an increased risk for triple negative breast cancer, which lacks expression of estrogen receptor (ER), progesterone receptor, and HER2 [9,10,11], while postmenopausal women with obesity have an increased risk for ER⁺ breast cancer [12,13]. Women with obesity tend to develop more aggressive breast cancers regardless of menopausal status [9,14,15,16]. Within breast tissue, obesity alters mammary epithelial cells to increase ER⁺ luminal cells, decrease basal/myoepithelial cells and enhance stem and progenitor activity [17]. This indicates obesity may predispose individuals to breast cancer by expanding the epithelial populations that give rise to the most common subtypes of breast cancer [18,19,20,21]. Additionally, obesity increases levels of adipokines and cytokines secreted by adipocytes, leading to enhanced macrophage recruitment and chronic inflammation in the mammary gland [17,22,23,24,25]. These studies suggest obesity changes the mammary microenvironment to create a supportive niche for tumorigenesis. Further, the changes to the mammary gland as a result of obesity may increase the susceptibility to chemically-induced carcinogenesis, however, more research is needed to investigate this hypothesis.

Acrylamide is classified as a probable carcinogen [26] and is produced as a biproduct of cooking starchy foods at high temperatures [27]. The obesity-inducing Western diet is made up of foods that have high acrylamide content [28,29], suggesting that individuals with obesity may be exposed to higher levels of acrylamide. Recommendations to reduce acrylamide production in food are provided by the United States Federal Drug Administration [30], and European companies selling foods containing acrylamide are required to report acrylamide levels due to its adverse health effects [31]. Recently, acrylamide exposure was found to enhance weight gain in mice [32] and zebrafish [33] as well as increase fat droplet accumulation in differentiated 3T3-L1 cultured adipocytes [32]. These studies suggest that acrylamide acts as an obesogen to enhance obesity in these models. Acrylamide is metabolized by the cytochrome P450 enzyme 2E1 (CYP2E1) to glycidamide or conjugated to glutathione for excretion [34,35]. Glycidamide is a genotoxic epoxide that forms DNA adducts at the N7 position of guanine and N3 position of adenine [36,37,38] and can increase tumor formation with long-term exposure in mice and rats [39]. Acrylamide exposure also leads to elevated levels of oxidative stress in a variety of tissues in vivo [40,41,42,43,44,45], although the mechanism for generation of oxidative stress has not been identified. Oxidative stress leads to the generation of reactive oxygen species (ROS) that can react with RNA and DNA to form adducts. Epidemiological studies suggest acrylamide exposure increases breast cancer risk in women [46,47,48,49], potentially through glycidamide-induced DNA damage. However, interactions of acrylamide with other breast cancer risk factors have not been explored.

Here we utilize a high-fat diet (HFD) mouse model of obesity to investigate the effects of obesity and acrylamide exposure on the mammary gland. We found that acrylamide treatment did not enhance obesity in lean or obese mice. However, obesity increased expression levels of CYP2E1 in the liver, which may elevate circulating levels of the carcinogenic metabolite, glycidamide. Mammary epithelial cells isolated from obese mice had enhanced DNA damage and oxidative stress following exposure to acrylamide. In COMMA-D epithelial cells, we observed that treatment with glycidamide increased DNA damage, while treatment with acrylamide promoted oxidative stress. Loss of CYP2E1 expression in COMMA-D cells rescued the acrylamide-induced increase in oxidative stress, suggesting that local metabolism of acrylamide to glycidamide elevates oxidative DNA damage in epithelial cells. These results suggest that acrylamide and obesity may interact to promote mammary epithelial DNA damage and oxidative stress via CYP2E1 to increase breast cancer risk.

2. Materials and Methods

2.1. Animal Studies

All animal procedures were approved by the University of Wisconsin Institutional Animal Care and Use Committee, per guidelines published by the NIH Guide for the Care and Use of Laboratory Animals (Protocol No. V005188). Female FVB/N mice (FVB/N-Tac) were purchased from Taconic Biosciences and housed in AAALAC-accredited facilities. Three-week-old FVB/N female mice were fed a low-fat (LFD; 16% kcal from fat, Teklad Global, 2020X) or HFD (60% kcal from fat, Test Diet #58Y1) and were provided with sterile filtered water or water supplemented with 0.7 mM acrylamide (Sigma-Aldrich, A9099) for 16 weeks. Food and water were provided ad libitum. Body weights were measured weekly. After 16 weeks, mice were injected with 200 mg/kg body weight of 5’-bromo-2’-deoxyuridine (BrdU; Fisher Scientific AC228590010) diluted with PBS one hour prior to euthanasia. All mice were in diestrus at time of euthanasia to control for cyclic changes in ovarian hormones. Euthanasia was performed by CO₂ asphyxiation. The right inguinal mammary gland was collected and fixed in 10% neutral buffered formalin (Epredia, 5701) for 48 hours then embedded in paraffin. The left inguinal and thoracic mammary glands were collected and digested for one hour at 37°C in DMEM (Corning, 10-017-CV) supplemented with 10% FBS (Gibco, A52567-01), 1% Antibiotic-Antimycotic Solution (Corning, 30-004-CI), 10 µg/mL insulin (Sigma-Aldrich, I0516), 5 ng/mL human epidermal growth factor (Sigma-Aldrich, E9644), 0.5 µg/mL hydrocortisone (Sigma-Aldrich, H0888), 3 mg/mL collagenase A (Sigma-Aldrich, 11088793001), and 100 U/mL hyaluronidase (Sigma-Aldrich, H3506). Digested tissue was cryopreserved for in vitro studies. The liver and visceral fat were collected, weighed, and a section of the liver was snap frozen for molecular analyses or fixed and embedded in paraffin.

The estrus cycle was tracked using vaginal cytology for 14 days prior to euthanasia. Vaginal epithelial cells were collected by washing the vaginal opening with 50 µL PBS and placed on microscope slides (Fisherbrand, 22-034-979). The slides were then stained with Wright’s Stain (Ricca Chemical Company, 9350-16), and the stage of the estrus cycle was assessed using light microscopy as described [50].

2.2. Mammary Gland Dissociation

Digested mammary tissues were plated for 1 hour in DMEM supplemented with 10% FBS, 1% Antibiotic-Antimycotic Solution, 10 µg/mL insulin, 5 ng/mL human epidermal growth factor, and 0.5 µg/mL hydrocortisone to separate the stromal vascular fraction and epithelial organoids. After one hour, the organoids were removed and pelleted. Cells were washed with 1X PBS, disrupted using a 20 g needle (Fisher Scientific, 14817209), and incubated with 2 mL of 0.25% trypsin/EDTA (Corning, 25-053-CI) for 10 minutes in a 37°C water bath. Trypsin was quenched with media containing FBS and 0.1 mg/mL DNase I (Sigma-Aldrich, 10104159001), filtered with a 40 µm cell strainer (Falcon, 352340), and counted with a hemocytometer (Hausser Scientific, 3200).

2.3. Cell Culture and Transfection Experiments

COMMA-D mammary epithelial cells were provided by Dr. Charlotte Kuperwasser (Tufts University, Boston, MA). 293T cells were obtained from ATCC (CRL-3216). All cells were cultured in DMEM supplemented with 10% FBS and 1% Antibiotic-Antimycotic Solution at 37°C at 5% CO₂. COMMA-D cells were treated with 9.8 µM acrylamide, 0.5 mM glycidamide, 100 µM hydrogen peroxide, or vehicle for 24 hours.

Bacterial stocks of Cyp2e1 shRNA (Millipore Sigma, TRCN0000011869) and human Cyp2e1 (DNASU Plasmid Repository, HsCD00942861) were grown for 2 days on agar with 25 mg/mL ampicillin (Fisher Scientific, BP90225) in a 37°C incubator. Single colonies were isolated and expanded for 8 hours in 1 mL LB medium, then 100 µL of that stock was further expanded overnight in 100 mL LB medium at 37°C in a shaker. Plasmids were extracted using the Invitrogen PureLink HiPure Plasmid Filter Maxiprep Kit (K210016). Three µg of isolated plasmid, 2 µg pCMV ΔR 8.2 Δvpr, and 1 µg pCMV-VSVG plasmids were transfected into 293T cells with TransIT-2020 (Mirus, MIR 5400). Transduced 293T cells were grown for 24 hours, then the media was removed and filtered with 0.45 µm syringe filters (Fisherbrand, 09-720-005) and incubated with COMMA-D cells. COMMA-D cells were selected with 0.04 mg puromycin (Fisher Scientific, AAJ67236XF) or 0.1 mg blasticidin (Thermofisher, R21001) for 5 days.

2.4. Immunohistochemistry and Immunofluorescence

Paraffin-embedded tissues were sectioned by the Experimental Pathology Laboratory (Carbone Cancer Center, University of Wisconsin-Madison) then stained with hematoxylin and eosin (H&E) to quantify mammary adipocyte diameters and lipid droplets in the liver. Tissues were stained as described [17]. Slides were blocked in 1% BSA in TBST, 5% fish gel in TBST, or using the MOM kit (Vector Labs, BMK-2202) for one hour. Primary antibodies were incubated overnight at 4°C and included cleaved caspase-3 (1:400; Cell Signaling, 9661T), BrdU (1:50; Accurate Chemical & Scientific Corp, OBT0030S), F4/80 (1:100; Biolegend, 123102), and 8-OHdG (1:100; Fisher Scientific, 501015749). The secondary antibodies were biotinylated goat anti-rabbit (1:500; Vector Labs, BA-1000), biotinylated rabbit anti-rat (1:400; Vector Labs, BA-4000), biotinylated horse anti-mouse (1:500; Vector Labs, 30044), Alexa Fluor 488 goat anti-rabbit H+L (1:250; Invitrogen, A11008), and Alexa Fluor 546 goat anti-rat H+L (1:250; Invitrogen, A11081). Biotinylated secondary antibodies were visualized with ImmPACT DAB Peroxidase Substrate Kit (Vector Labs, SK-4105). For immunofluorescent stains, nuclei were counterstained with DAPI. All tissues were imaged on the Nikon Eclipse E600. All images were quantified using ImageJ (NIH). Five ducts per tissue section were imaged from five mice per group. Positive and negative cells were counted in the ducts and the number of positive cells were divided by the total number of cells and multiplied by 100. F4/80 staining was quantified by taking five images per tissue section from five mice per group, and the number of crown-like structures within the field of view was quantified.

2.5. Immunocytochemistry

COMMA-D cells were plated on 8-well chamber slides (Falcon, 354118), treated for 24 hours after reaching confluence, and fixed with 100% ice cold methanol for 10 minutes at -20°C. Cells were permeabilized with 0.1% Triton-X in PBS. The slides were either blocked in 1% BSA in TBST for one hour at room temperature or incubated with 2M HCl for one hour at room temperature prior to labeling for BrdU. Primary antibodies included 8-OHG (1:200; NOVUS, NB600-1508), γH2AX (1:100), and BrdU (1:50). The secondary antibodies included Alexa Fluor 546 donkey anti-goat H+L (Invitrogen, A11056), Alexa Fluor 488 goat anti-rabbit H+L, and Alexa Fluor 546 goat anti-rat H+L. Nuclei were counterstained with DAPI and imaged with the Nikon Eclipse E600. Three experiments were plated in duplicate, and five images per well were taken. The mean cell intensity was quantified using ImageJ.

2.6. Alkaline Comet and FLARE Assays

Alkaline comet assays were performed in triplicate following the Trevigen CometAssay HT (4252-040-K) protocol with some modifications. Five hundred mammary epithelial cells or COMMA-D cells were combined with 50 µL of CometAssay LMAgarose (R&D Systems, 4250-200-03) and spread on CometSlides (R&D Systems, 4250-200-03) pre-warmed to 37°C. After the agarose gels solidified for 30 minutes at 4°C, slides were immersed in lysis buffer (R&D Systems, 4250-050-K) overnight at 4°C. Slides were then transferred to Alkaline Unwinding Solution (200 mM NaOH, 1 mM EDTA) for one hour at room temperature in the dark. Gel electrophoresis was performed in Alkaline Electrophoresis Solution (300 mM NaOH, 1 mM EDTA) for 40 minutes at 25 V at 4°C. Cells were stained with 100 µL SYBR Gold Nucleic Acid Gel Stain (Invitrogen, S11494) for 30 minutes at room temperature. Fifty to sixty cells from each slide were imaged.

Fragment Length Analysis using Repair Enzymes (FLARE) assays with the formamidopyrimidine DNA glycosylase (FPG) enzyme were performed in triplicate under neutral conditions following the Trevigen CometAssay HT (4252-040-K) protocol with some modifications. COMMA-D cells were plated and lysed overnight as described above. Slides were then transferred to Enzyme Buffer (40 mM HEPES, 0.5 mM EDTA, 0.1 M KCl, pH7.6) for one hour at 4°C. FPG was added at 1:1000 to pre-warmed Enzyme Buffer with 0.2 mg/mL BSA. 100 µL of FPG Enzyme Buffer was added to each sample gel and incubated for one hour at 37°C in a humidified chamber. Control gels received Enzyme Buffer with BSA without FPG. Slides were transferred to 1X Neutral Electrophoresis Buffer (0.05 M Tris Base, 0.15 M Sodium Acetate) for 30 minutes at 4°C then gel electrophoresis was performed in 1X Neutral Electrophoresis Buffer at 25 V for 40 minutes at 4°C. After incubation in DNA Precipitation Solution (1 M Ammonium Acetate in 95% EtOH), fixation with 70% EtOH, and drying for 15 minutes at 37°C, cells were stained with SYBR Gold. Twenty-five to thirty-five cells from each slide were imaged. All slides from the alkaline comet assays and FLARE assays were imaged with Nikon Eclipse E600 and analyzed on ImageJ with OpenComet v1.3.1.

2.7. ROS Assay

Reactive oxygen species were measured following the protocol provided by CellROX Green Reagent Kit (Thermofisher, C10444). Briefly, COMMA-D cells were cultured on a 96-well plate until confluence, then treated. After 24 hours, cells were washed with 1X PBS and incubated with 5 µM/well CellROX Green for 30 minutes at 37°C. Wells were washed three times with PBS, fixed in 3.7% formaldehyde, and nuclei were counterstained with DAPI. Cells were imaged with the Keyence BZ-X710, and images were quantified by measuring fluorescent intensity divided by DAPI intensity using ImageJ.

2.8. Western Blots

Protein was extracted from tissues and COMMA-D cells using RIPA buffer and Protease Inhibitor Cocktail (Promega, G651). Isolated protein was loaded into 10% SDS Page gels (Bio-Rad, 4568034) at 20 µg/well. Electrophoresis was performed at 150 V in 10% Tris/Glycine/SDS Buffer (Bio-Rad, 1610732). Protein was transferred from the gel to a nitrocellulose membrane (GE Healthcare, RPN303D) in 10% Tris/Glycine Buffer (Bio-Rad, 1610771) at 100 V for one hour. Protein transfer was confirmed with Ponceau S staining (Sigma, P7170). The nitrocellulose membrane was washed with 0.1% TBST until Ponceau staining cleared and was blocked with 5% milk in 0.1% TBST for one hour at room temperature on an orbital shaker. Membranes were then probed with either anti-CYP2E1 (1:500; Thermo Scientific, 19937-1-AP) or anti-GAPDH (1:10,000; Invitrogen, MA5-15738) overnight at 4°C on an orbital shaker. After washing with 0.1% TBST, membranes were probed with secondary goat anti-rabbit (1:10,000; Invitrogen, 31460) or anti-mouse (1:10,000; Invitrogen, 31430) conjugated to HRP for one hour at room temperature on an orbital shaker. Membranes were treated with SuperSignal West Femto Maximum Sensitivity Substrate (Thermo Scientific, 34095) for five minutes then the signal was detected using film (GeneMate, F-9023) developed on the All-Pro Imaging Corp 100 Plus Automatic X-Ray Film Processor. Protein content was quantified by measuring pixels on ImageJ.

2.9. Quantitative Real Time Polymerase Chain Reaction

RNA was extracted with TRIzol (Life Technologies, 15596018) and PureLink RNA Mini Kit (Invitrogen, 12183018A) or with the Quick-DNA/RNA Miniprep Kit (Zymo Research, D7001). The RNA was reverse transcribed using the High-Capacity cDNA Reverse Transcription Kit (Applied Biosystems, 4368814) and Biometra Thermal Cycler (Analytik Jena). Quantitative PCR was performed with iTaq Universal SYBR Green Supermix (Bio-Rad, 1725121) on a Bio-Rad CFX Connect Real-Time PCR Detection System (Bio-Rad). Data were analyzed using the ΔCT or the ΔΔCT method. Transcripts were normalized to hypoxanthine phosphoribosyl transferase (HPRT). Primer sequences are found in Table S1.

2.10. Glutathione Assay

Glutathione and oxidized glutathione dimers (GSSG) were measured using the Glutathione Colorimetric Detection kit (Thermo Scientific, EIAGSHC). COMMA-D cells were lysed in 5% 5-sulfosalicylic acid dihydrate, and samples were diluted 1:5 in assay buffer. GSSG was measured by treating samples and standards with 2-vinylpyridine. Absorbance was read at 405 nm, and cellular concentrations of total glutathione and GSSG were calculated from the respective standard curves.

2.11. Statistical Analysis

Results are reported as the mean ± standard error of the mean (s.e.m.). Statistical differences were determined using one-way analysis of variance (ANOVA) and Tukey’s multiple comparisons posttest, unless otherwise noted. A p-value of ≤0.05 denotes significant value. All statistical analyses were performed with GraphPad Prism 9.4.1 (GraphPad Software).

3. Results

3.1. Acrylamide Does Not Enhance Obesity-Associated Changes to the Mammary Gland

To examine how acrylamide impacts the mammary gland under conditions of obesity, FVB/N female mice received LFD or HFD and water or 0.7 mM acrylamide-supplemented water ad libitum for 16 weeks. The dose of acrylamide administered was similar to acrylamide exposure levels in humans [48], and chronic treatment of mice and rats with 0.7 mM acrylamide was found to promote mammary tumorigenesis [51]. After 16 weeks, mice fed HFD were significantly heavier than mice fed LFD (p=0.001, Figure 1A). However, acrylamide treatment had no impact on final body weights (Figure 1A) or weight gain (Figure S1A) compared to the respective controls. In addition to body weights, HFD-fed mice had significantly increased visceral fat weights (p=0.0001, Figure 1B) and mammary gland weights (p=0.02, Figure 1C), but acrylamide-treated mice did not significantly differ in either endpoint compared to vehicle-treated mice on each respective diet (Figure 1B, C). Consistent with elevated in mammary gland mass, HFD-fed mice had increased adipocyte diameters compared to LFD-fed mice (p<0.0001, Figure 1D). Acrylamide treatment of LFD-fed mice did not alter adipocyte diameters compared to control LFD-fed mice (Figure 1D). In contrast, acrylamide treatment of HFD-fed mice had a further significant increase in adipocyte diameters compared to HFD-fed mice (p<0.0001, Figure 1D). A hallmark of obesity is the formation of crown-like structures which are formed when F4/80⁺ macrophages are recruited into the mammary gland and surround dying adipocytes [52,53]. F4/80⁺ crown-like structures were significantly elevated in mammary glands of HFD-fed mice compared to LFD-fed mice (p=0.01, Figure 1E). However, acrylamide treatment did not further enhance crown-like structures in the mammary glands compared to their respective controls (Figure 1E). Taken together these results show that HFD-induced obesity and the addition of acrylamide treatment further increased adipocyte diameters, but acrylamide did not significantly increase weight gain or inflammation.

3.2. Acrylamide Induces Systemic Changes in CYP2E1 Expression and DNA Damage

The liver is a site of ectopic lipid deposition in obesity, and we hypothesized that acrylamide exposure could enhance lipid in this tissue. Liver weights were not significantly different in HFD-fed mice compared to LFD-fed mice, and acrylamide treatment did not enhance liver weights compared to controls (Figure S1B). Although not severe enough to impact liver weight, HFD-fed mice had significantly more fat droplets in the liver than LFD-fed mice (p<0.0001, Figure 2A). However, acrylamide-treatment did not increase lipid droplets in the liver of either obese or lean mice, suggesting acrylamide does not enhance fat storage in the liver to elevate the negative effects of obesity.

Mammary gland development and function is regulated by the ovarian steroids estrogen and progesterone [54]. To examine how exposure to HFD or acrylamide altered ovarian function, vaginal cytology was performed for two weeks to identify changes in the estrus cycle. As we observed previously [17], exposure to HFD did not significantly alter time spent in either estrus or diestrus compared to LFD-fed mice (Figure S1C, D). Similarly, acrylamide-treated mice did not differ in the number of days spent in estrus or diestrus compared to control vehicle-treated mice (Figure S1C, D).

Recent studies suggest that individuals with obesity have enhanced activity of the acrylamide-metabolizing P450 enzyme, CYP2E1, in the liver [55]. Relative expression of Cyp2e1 in the liver was not significantly different among mice in any of the groups, although levels were mildly elevated in the obese, acrylamide-treated group (Figure S1E). However, HFD-fed mice supplemented with acrylamide had significantly higher levels of CYP2E1 protein in the liver compared to vehicle-treated LFD-fed mice (p=0.04, Figure 2B). These results suggest that obesity and acrylamide together increase the levels of CYP2E1 protein in the liver. Elevated levels of CYP2E1 may increase metabolism of acrylamide to the genotoxic metabolite, glycidamide. To test this hypothesis, we examined DNA damage in peripheral blood mononuclear cells (PBMC) using alkaline comet assays. No difference in DNA damage was observed in PBMC isolated from LFD and HFD-fed mice (Figure 2C). However, PBMCs isolated from both LFD and HFD-fed acrylamide-treated mice demonstrated elevated DNA damage compared to both groups of control mice (Figure 2C). The DNA damage present in circulating cells suggests that exposure to acrylamide can induce systemic DNA damage, potentially via glycidamide.

3.3. Acrylamide Exacerbates DNA Damage and Apoptosis in Mammary Glands from Obese Mice

Both acrylamide and glycidamide are detectable in the rat mammary gland within two hours of dosing [56], indicating mammary epithelial cells are exposed to both acrylamide and its metabolite glycidamide. To test how long-term exposure to acrylamide impacts DNA damage in mammary epithelial cells, alkaline comet assays were performed on mammary epithelial cells isolated from mice from each group. HFD-fed mice had significantly elevated DNA damage compared to LFD-fed mice (p<0.0001, Figure 3A). LFD-fed mice treated with acrylamide also had significantly more DNA damage compared to lean mice alone (p=0.004, Figure 3A), with comparable levels of DNA damage as control HFD-fed mice. Acrylamide-treated HFD-fed mice had significantly more DNA damage compared to all three other groups (Figure 3A), suggesting that obesity and acrylamide together enhance mammary epithelial DNA damage.

Oxidative DNA adducts are a form of DNA damage that can promote DNA-protein crosslinks, stall transcription and replication, and act as a mutagen [57]. 8-hydroxy-2' -deoxyguanosine (8-OHdG) is one of the most common oxidative adducts and is highly mutagenic [58]. Mammary ducts from HFD-fed mice had significantly increased 8-OHdG⁺ cells compared to LFD-fed mice (p=0.04, Figure 3B). LFD-fed mice treated with acrylamide also had significantly more 8-OHdG⁺ cells compared to LFD-fed mice without acrylamide treatment (p=0.009, Figure 3B). Mammary ducts from HFD-fed, acrylamide-treated mice had the highest number of 8-OHdG⁺ epithelial cells compared to all other groups (Figure 3B). These results suggest that obesity and acrylamide enhance oxidative DNA damage in mammary epithelial cells.

Due to the enhanced oxidative DNA damage, we next analyzed how antioxidant levels were impacted by acrylamide treatment and diet. Levels of catalase (Cat) and nitric oxide synthase 2 (Nos2) were mildly elevated in HFD-fed, acrylamide-treated mice (Figure S2A, B). Levels of superoxide dismutase (Sod1) did not differ across treatment groups (Figure S2C). These data suggest that acrylamide exposure did not significantly alter antioxidant expression in mammary epithelial cells.

As elevated levels of DNA damage can lead to increased apoptosis [59], we examined cleaved caspase-3 in mammary epithelial cells. HFD-fed mice had enhanced levels of cleaved caspase-3 compared to LFD-fed mice (p=0.01, Figure 3C). Additionally, epithelial cells from LFD-fed acrylamide-treated mice had elevated cleaved caspase-3 compared to lean, vehicle-treated mice (p=0.03, Figure 3C). HFD-fed, acrylamide-treated mice had the most cleaved caspase-3⁺ epithelial cells compared to all other groups (Figure 3C). These results suggest that obesity and acrylamide exposure enhance mammary epithelial DNA damage and apoptosis.

3.4. Glycidamide Increases Single and Double Strand DNA Breaks, but Acrylamide Enhances Oxidative Stress and Oxidative DNA Damage in COMMA-D Cells

To understand the mechanism of DNA damage in mammary epithelial cells due to acrylamide exposure, COMMA-D epithelial cells were cultured and treated with 9.8 µM acrylamide, 0.5 mM glycidamide, or vehicle for 24 hours. COMMA-D cells were also treated with 100 µM hydrogen peroxide as a positive control for DNA damage. Hydrogen peroxide treatment enhanced levels of DNA strand breaks measured by alkaline comet assays (p<0.0001, Figure S3A). When COMMA-D cells were treated with acrylamide, we did not observe elevated DNA strand breaks measured by the comet assays (Figure 4A). However, glycidamide treatment significantly increased DNA damage within the cells compared to both vehicle and acrylamide treatment (Figure 4A). Hydrogen peroxide treatment also enhanced expression of γH2AX (p=0.002, Figure S3B). In contrast, γH2AX levels were not different between vehicle and acrylamide treatment (Figure 4B). Glycidamide treatment increased γH2AX levels compared to vehicle-treated cells (p=0.04, Figure 4B). These results suggest glycidamide, rather than acrylamide, increases DNA strand breaks.

To assess if acrylamide or glycidamide enhances oxidative stress in COMMA-D cells, levels of intracellular ROS were measured. Hydrogen peroxide treatment enhanced ROS (p<0.0001, Figure S3C). Surprisingly, ROS levels were significantly higher in acrylamide treated cells than either vehicle or glycidamide-treated cells (Figure 4C). Elevated oxidative stress can also result in the formation of oxidative adducts on RNA in addition to adducts on DNA. We examined treated COMMA-D cells for the oxidative RNA adduct 8-hydroxyguanine (8-OHG). We observed that hydrogen peroxide treatment did not significantly increase 8-OHG levels compared to vehicle (Figure S3D). However, acrylamide treatment significantly increased 8-OHG adducts compared to both vehicle and glycidamide treatment (Figure 4D), suggesting that acrylamide elevates levels of oxidative stress. Together, this suggests acrylamide enhances oxidative stress via increased ROS levels within cells led to damaged cytosolic RNA.

Glutathione is one of the main antioxidants involved in cellular ROS detoxification. Glutathione is oxidized in response to ROS or can be directly conjugated to acrylamide to promote excretion of acrylamide. Therefore, we predicted glutathione and oxidized glutathione levels would be depleted in acrylamide-treated cells. Surprisingly, glutathione and oxidized glutathione levels were elevated in glycidamide-treated cells compared to both vehicle and acrylamide-treated cells (Figure 4E). However, acrylamide treatment led to elevated expression levels of other antioxidants, specifically Cat and Nos2 but not Sod1 (Figure 4F). Overall, this data indicates that glycidamide enhanced DNA strand breaks, but acrylamide treatment elevated oxidative stress and oxidative RNA damage in COMMA-D cells.

3.5. Knockdown of CYP2E1 in COMMA-D Cells Rescues Acrylamide-Induced Oxidative Stress and Oxidative DNA Damage

Activation of CYP2E1 has been shown to increase cellular oxidative stress [60,61]. While CYP2E1 is highly expressed in the liver (Figure S4A), we observed that CYP2E1 protein was also present in COMMA-D cells (Figure S4A), and Cyp2e1 transcripts were detected in primary mammary epithelial cells isolated from mice in all treatment groups (Figure S4B). We hypothesized that acrylamide-associated oxidative stress may be due to activation of CYP2E1 within epithelial cells. COMMA-D cells were transduced with lentivirus encoding either shRNA scrambled control (shScram) or shRNA targeting Cyp2e1 (shCyp2e1) as well as lentivirus to overexpress Cyp2e1 to generate stable cell lines. Compared to shScram cells, shCyp2e1 cells had significantly reduced CYP2E1 protein (p=0.03, Figure 5A), while overexpressing cells had elevated CYP2E1 expression compared to controls (p=0.006, Figure S4C). When shScram cells were treated with acrylamide, we

We also examined levels of oxidative RNA adduct 8-OHG. In shScram cells, acrylamide treatment significantly increased 8-OHG RNA adducts compared to shScram cells treated with vehicle and glycidamide (Figure 5C). However, treatment of shCyp2e1 cells with acrylamide did not enhance 8-OHG adducts compared to vehicle or glycidamide treated cells (Figure 5C). Further, overexpression of CYP2E1 resulted in enhanced levels of 8-OHG adducts after acrylamide treatment compared to all treatments of control cells and overexpressing CYP2E1 cells treated with vehicle or glycidamide (Figure S4E). These results support the hypothesis that acrylamide-driven oxidative adducts are a result of CYP2E1 activation.

Oxidative DNA damage can also be assessed utilizing a modified comet assay called the fragment length analysis using restriction enzymes (FLARE) assay. This assay includes treatment of permeabilized cells with FPG enzyme, which excises oxidized DNA bases to create single strand DNA breaks detectable by electrophoresis. These DNA breaks would otherwise go undetected by an unmodified comet assay. We hypothesized that acrylamide treatment increased oxidative DNA damage due to elevated oxidative stress. In shScram cells, we observed increased DNA damage in vehicle-treated cells with FPG (p=0.04, Figure 5D), indicating basal oxidative DNA damage in these cells. Consistent with our previous observation in comet assays (Figure 4A), no differences in DNA damage were detected in shScram cells treated with acrylamide without FPG compared to vehicle treated cells without FPG (Figure 5D). However, FPG treatment revealed that acrylamide-treated shScram cells had significantly increased DNA damage compared to non-FPG, acrylamide-treated cells (p<0.0001) as well as cells that were treated with vehicle and vehicle with FPG (Figure 5D). When shScram cells were treated with glycidamide, we observed increased DNA damage compared to vehicle-treated cells (Figure 5E). However, treatment of cells with FPG did not further enhance the levels of DNA damage (Figure 5E). These data indicate that glycidamide treatment does not induce oxidative DNA damage.

We also examined how loss of CYP2E1 expression impacted oxidative DNA damage measured with FLARE assays. Similar to shScram cells (Figure 5D), vehicle-treated shCyp2e1 cells showed elevated DNA damage when treated with FPG compared to without FPG (p=0.001, Figure 5F). When shCyp2e1 cells were treated with acrylamide, DNA damage was not different compared to vehicle treated cells either with or without the addition of FPG (Figure 5F), showing that loss of CYP2E1 eliminated oxidative DNA damage due to acrylamide metabolism. In contrast, glycidamide-treated shCyp2e1 cells had significantly increased DNA damage compared to shCyp2e1 vehicle-treated cells with and without FPG (p<0.0001, Figure 5G), but the addition of FPG to glycidamide-treated cells did not further increase DNA damage (Figure 5G). Overall, this data suggests acrylamide but not glycidamide enhances oxidative DNA damage mediated by CYP2E1.

4. Discussion

Obesity and carcinogen exposure both increase the risk for breast cancer [3,4,12,13], but there is limited knowledge on how obesity and carcinogens interact to promote breast cancer risk. Acrylamide has been suggested to be both a carcinogen and an obesity-inducing agent [26,32,33], so we addressed the knowledge gap on the effects of acrylamide exposure on weight gain and DNA damage in mammary epithelial cells under conditions of obesity. Our results suggest that exposure to 0.7 mM of acrylamide has limited effects as an obesity-inducing agent but enhances mammary epithelial DNA damage, which may increase the risk for breast cancer. These results are significant due to the high prevalence of acrylamide in the Western diet [28,29]. While CYP2E1 is expressed within the liver, we also observed that CYP2E1 is expressed in mammary epithelial cells, suggesting that acrylamide is also metabolized within the mammary gland. Metabolism of acrylamide by CYP2E1 led to elevated epithelial oxidative stress and increased DNA damage. Elevated levels of acrylamide consumption through the Western diet could promote genotoxic effects through both ROS generation and glycidamide exposure within mammary epithelial cells. Consistent with studies focused on patients with obesity [55], obese mice had elevated levels of CYP2E1. Increased obesity through the Western diet may further fuel DNA damage through elevated expression of CYP2E1 to metabolize acrylamide to genotoxic intermediates. Additional studies are necessary to identify how obesity and acrylamide exposure alter mammary tumorigenesis.

An obesogen is a compound that can cause both metabolic dysfunction and weight gain [62]. While acrylamide has been shown to enhance weight gain in mouse and zebrafish models [32,33], there is contradictory epidemiological evidence associating acrylamide exposure with elevated weights in humans [63,64,65]. We found that treatment of mice with 0.7 mM of acrylamide-supplemented water did not increase body, visceral fat, liver, or mammary gland weights after exposure for 16 weeks in lean or obese mice. These results are in agreement with another study where mice and rats were exposed to acrylamide at the same dose for two years and did not identify changes in body weight [51]. However, Lee et al. reported increased body weight with acrylamide treatment when 50 µg/kg of acrylamide was administered to mice through oral gavage [32]. Acrylamide is rapidly absorbed through the gut and has an elimination half-life of 1-6 hours in mice and rats, depending on the tissue type [66,67,68]. It is possible that administration of a large bolus of acrylamide through oral gavage could have different effects on weight gain than continuous administration of lower levels of acrylamide in drinking water. Other studies have demonstrated that different doses of acrylamide did not impact body weight in rats but instead increased serum levels of cholesterol, glucose, and triglycerides [69,70,71]. Elevated levels of cholesterol, triglycerides, and glucose are associated with heart disease and diabetes [72,73], suggesting that acrylamide could act as an obesogen to increase the risk for metabolic diseases. A limitation of our study is that we did not measure metabolic markers, such as serum triglycerides, in addition to weight gain to investigate both the weight-inducing and metabolic properties of acrylamide. We did find increased adipocyte diameters in obese, acrylamide-treated mice which may indicate acrylamide-associated disruption to lipid metabolism, but more research is needed to elucidate how acrylamide impacts adipocyte metabolism in this model.

The acrylamide metabolite, glycidamide, is thought to be the agent responsible for DNA damage and carcinogenesis after acrylamide exposure rather than acrylamide itself. Glycidamide is genotoxic and forms mutagenic adducts at a faster rate than acrylamide [74]. We found glycidamide, rather than acrylamide, induced DNA strand breaks in COMMA-D epithelial cells assessed with comet assays, consistent with literature demonstrating glycidamide induces DNA adducts and subsequent DNA damage [75,76,77]. Acrylamide can form guanine adducts but they form at a slower rate than glycidamide adducts [74], do not undergo spontaneous depurination to become mutagenic [74,78], and therefore have fewer biological consequences [74]. We observed elevated DNA damage in epithelial cells from obese, acrylamide-treated mice, compared to all other groups. Elevated levels of CYP2E1 protein in the liver likely lead to increased conversion of acrylamide to glycidamide and higher circulating levels of mutagen glycidamide. In contrast, isolated PBMCs from lean and obese mice showed similar levels of DNA damage from acrylamide exposure. This may be in part due to the short lifespan of PBMCs [79]. Additionally, increased DNA damage can induce apoptosis [59], which can prevent the analysis of cells with extreme DNA damage and identification of differences across these groups. The glycidamide-induced DNA damage observed in epithelial cells may elevate the risk for cancer. B6C3F₁ mice treated with glycidamide developed significantly more hepatic mutations and hepatocarcinomas than vehicle and acrylamide-treated mice [37]. Further, in mouse embryo fibroblasts with a human-TP53 knock-in gene, treatment with glycidamide created a mutational signature characterized by A>T and T>A point mutations [80,81]. This mutational signature has been found in breast cancer patients and individuals who smoke [80,81]. In contrast acrylamide treatment of these cells did not alter the amount or types of mutations from standard cell culture conditions [80,81]. However, these mouse embryo fibroblasts had low to no expression of CYP2E1 [80,81], preventing the analysis of oxidative mutational signatures. Future work is necessary to understand how obesity impacts the mutational signatures of glycidamide and acrylamide-associated oxidative mutations and subsequent mammary tumorigenesis.

A variety of antioxidant pathways may be activated in response to cellular oxidative stress. Marković Filipović et al. demonstrated acrylamide treatment promoted the activation of iNOS, SOD1, and SOD2 antioxidants [82], however we did not observe significantly elevated expression of antioxidants Cat, Nos2, or Sod1 in mammary epithelial cells with acrylamide treatment. Recent work has shown that basal and luminal epithelial cell populations in the mammary gland have different antioxidant capacities [83]. Basal mammary epithelial cells primarily utilize glutathione-dependent mechanisms for antioxidant control, while luminal cells utilize both glutathione-dependent and independent pathways for antioxidants, including SOD. We did not separate basal and luminal cells during RNA extraction, potentially masking antioxidant responses in specific types of epithelial cells. We were surprised to detect increased levels of glutathione and oxidized glutathione in COMMA-D cells treated with glycidamide. This may indicate a role for glutathione-mediated metabolism of glycidamide in mammary epithelial cells. Glycidamide-glutathione conjugates have been identified in serum of rats [84], and mercapturic acid derivates of glycidamide-glutathione conjugates have been characterized in urine of humans [85], however little data has been reported on this conjugate in tissues. If glycidamide is conjugated to glutathione for elimination, we would expect glutathione levels to be reduced instead of elevated in COMMA-D cells. However, MCF7 breast cancer cells and CaCo-2 colon cancer cells treated with low concentrations of glycidamide led to depleted glutathione levels, while at a higher dose, glycidamide significantly increased glutathione levels in MCF7 cells and rescued the depleted glutathione levels in CaCo-2 cells [86]. These results suggest different doses of glycidamide have divergent effects on glutathione levels. Acrylamide can also be conjugated to glutathione for removal and excretion [35], indicating glutathione levels could also be depleted after acrylamide treatment [40,41,42,43,44]. Liver, brain, and kidney tissues had depleted glutathione after acrylamide treatment in vivo [40,41,42,43,44]. However, we did not observe any differences in glutathione or oxidized glutathione levels in acrylamide-treated cells compared to vehicle-treated cells. The half-life of the acrylamide-glutathione conjugate is approximately one hour in serum [84], indicating changes in glutathione levels may not have been captured at the 24-hour timepoint for in vitro acrylamide treatment. Additionally, serum acrylamide-glutathione conjugates have a shorter half-life than glycidamide-glutathione conjugates [84]. These studies suggest that the lack of change in glutathione levels in acrylamide-treated cells may be due to more efficient removal of acrylamide than glycidamide in COMMA-D cells.

We observed that acrylamide treatment enhanced oxidative DNA damage in COMMA-D cells, showing that DNA damage can occur through exposure to both glycidamide and acrylamide. Acrylamide metabolizing enzyme CYP2E1 utilizes NADPH and H⁺ to reduce oxygen to hydrogen peroxide and O₂^- [87], which promotes oxidative stress and mitochondrial dysfunction [88]. Loss of CYP2E1 in COMMA-D cells reversed acrylamide-induced ROS and oxidative DNA damage detected by FLARE assays, demonstrating that CYP2E1 metabolism of acrylamide was a source of DNA damage following acrylamide exposure. CYP2E1 is also induced by alcohol consumption [89], which is another risk factor for breast cancer [3,4]. In fatty liver disease, oxidative damage to the liver is thought to arise in part from CYP2E1-mediated oxidative stress [90], which is elevated with obesity and excess alcohol consumption [55,91,92,93,94]. Further, alcohol consumption under conditions of obesity worsens the severity of fatty liver disease through CYP2E1 activation in rodent models [95,96,97] and in humans [98]. These studies imply that the breast cancer risk factors of obesity and alcohol intake could interact through CYP2E1-associated oxidative stress to promote tumorigenesis. Genetics may also play a role in how CYP2E1 contributes to oxidative stress and DNA damage. Activating polymorphisms such as the c2 allele, which contains a cytosine to thymine and a guanine to cytosine transversion before the regulatory region of the CYP2E1 gene, have been shown to enhance the severity of fatty liver disease [92]. Further, an insertion polymorphism in the regulator region of the CYP2E1 gene increases the activity of CYP2E1 protein only under conditions of obesity [99], suggesting that both genotype and environment interact to increase risk for diseases associated with CYP2E1 activation. The addition of CYP2E1 substrates like alcohol and acrylamide may further contribute to disease progression. These highlight the importance and complexity of investigating the interaction between multiple risk factors to identify genetic and environmental mechanisms that promote breast cancer.

5. Conclusions

Overall, the data presented here demonstrate acrylamide and obesity interact to promote DNA damage and CYP2E1-mediated oxidative stress in mammary epithelial cells. These lesions may elevate the risk for breast cancer through mutagenesis from glycidamide and oxidative stress. Future studies may focus on CYP2E1 inhibitors in vivo to examine how CYP2E1 influences the combined effects of obesity and acrylamide exposure on mammary epithelial DNA damage and oxidative stress. CYP2E1 may be a target for future breast cancer therapies. This work highlights the need for prevention strategies to minimize and eliminate acrylamide from the diet as well as identify obesity interventions.